Operating truth
Layer 1What was sold, paid, shipped, refunded, delivered, or canceled?
Typical sources: Commerce · payments · WMS · carrier · helpdesk
[ WORKFLOWSY OPERATOR FIELD GUIDE ]2026 edition
The DTC Metrics Operating System.
A practical field guide to the numbers every direct-to-consumer company should define, collect, reconcile, and review—before building another dashboard.
174+
Metric definitions
10
Operating systems
4
Review cadences
Operator field guideWorkflowsy / Ecommerce
One shared definition for every decision.
The dictionary, operating contract, scorecard, metric traps, and 90-day roadmap for a DTC measurement layer your team can trust.
1Define the business question
2Reconcile the truth layers
3Assign the action owner
Workflowsy17 pages
[ THE OPERATING PRINCIPLE ]
A metric is not operational until the team agrees on what it means.
Lock the formula, population, timing, source, owner, freshness expectation, and action threshold before publishing it.
Marketing truth
Layer 2What does each channel claim credit for under its attribution settings?
Typical sources: Ad platforms · GA4 · attribution provider
Financial truth
Layer 3What does the business recognize after policy, timing, and cost treatment?
Typical sources: Accounting · ERP · finance model
[ THE EXECUTIVE SCORECARD ]
Leadership should not review every metric every day.
Start with twelve signals. Use the wider metric dictionary only when you need to diagnose movement.
Daily · weekly · monthly
Executive metricLensDefault review
01Net product sales
LensDemand
ReviewDaily / Weekly
02Revenue growth vs plan and prior year
LensDemand
ReviewWeekly
03Gross margin dollars and percent
LensEconomics
ReviewWeekly / Monthly
04Contribution margin dollars and percent
LensEconomics
ReviewWeekly / Monthly
05New customers and new-customer revenue
LensAcquisition
ReviewDaily / Weekly
06Blended CAC and paid-media CAC
LensAcquisition
ReviewDaily / Weekly
07MER and new-customer ROAS
LensMarketing
ReviewDaily / Weekly
08Store conversion rate and revenue per session
LensFunnel
ReviewDaily / Weekly
09AOV and units per transaction
LensBasket
ReviewDaily / Weekly
10Repeat purchase and cohort retention
LensRetention
ReviewMonthly
11Inventory days, stockouts, and aged inventory
LensOperations
ReviewWeekly
12Cash balance and forecast variance
LensFinance
ReviewWeekly / Monthly
[ A 90-DAY IMPLEMENTATION ROADMAP ]
Definitions before automation. Automation before sophisticated modeling.
The sequence is part of the system. Build trust in the control totals first, then add depth, then automate recurring decisions.
Days 0–301
Define and reconcile
- Choose one commerce control total for sales, orders, customers, discounts, and reversals.
- Publish the 12-metric scorecard with source, owner, freshness, and comparison period.
- Reconcile commerce sales to payments and financial revenue.
- Create a metric catalog with formula, grain, exclusions, timezone, source, SLA, and owner.
Days 31–602
Add customer, product, and lifecycle depth
- Create a canonical customer identity and explicit first-order policy.
- Build acquisition cohorts and mature retention and LTV views.
- Add SKU margin, returns, inventory, and stockout diagnostics.
- Connect CRM, subscriptions, helpdesk, 3PL, carrier, and payments to shared keys.
Days 61–903
Automate decisions
- Separate platform claims, blended business performance, and incrementality evidence.
- Automate anomaly detection and narrative reporting with human approval gates.
- Build forecast-versus-actual bridges for revenue, margin, spend, inventory, and cash.
- Retire duplicate dashboards only after reconciliation is stable and users accept the replacement.
[ SEVEN METRIC TRAPS ]
The dashboard can be polished and still be wrong.
Most metric failures are contract failures: the population changed, the time window drifted, the source went stale, or nobody owned the action.
Calling platform ROAS revenue truth
Use platform attribution to optimize, then reconcile it to commerce and finance truth.
Using total profiles as an active audience
Separate marketable, reachable, and recently engaged profiles.
Mixing cohorts with different observation windows
Show eligible cohort dates and mark immature periods unavailable.
Treating returns and refunds as the same event
Keep sales reversals, refunds, and physical returns distinct.
Comparing retention on different populations
Use the same eligible months and original population before comparing.
Letting dashboards silently use stale data
Display source freshness and fail closed when critical inputs are outside SLA.
Building metrics without an action
Give every KPI an owner, cadence, threshold, and diagnostic path.
[ KEEP THE COMPLETE SYSTEM ]
Download the complete 17-page field guide.
Use it to audit an existing dashboard, create a metric catalog, or scope the foundation behind a new commerce intelligence system.
174+ metric definitions
12-metric executive scorecard
Ten operating domains
Four review cadences
Metric specification template
90-day implementation roadmap
Educational use only. Align every definition with your accounting policy, platform configuration, and data contracts.
[ FROM CATALOG TO OPERATING SYSTEM ]
The guide defines the contract. The build makes it run.
Use the resource yourself, or bring Workflowsy in where definitions, systems, and decisions have to meet.
Define
Choose the metrics and operating contracts that matter.
Start with an audit when the question, priority, or first build is still unclear.
Reconcile
Connect the systems and establish trusted control totals.
Build the data foundation when identity, timing, and source contracts are the constraint.
Automate
Turn evidence into recurring answers and action.
Launch Commerce Intelligence when the team needs cross-system monitoring, explanations, and workflows.
[ ONE USEFUL QUESTION ]
Which metric keeps starting an argument instead of a decision?
Bring the question to the Commerce Question Architect. It will map the evidence and definitions behind the first useful answer.